Color space conversion in MATLAB - Time & Space Complexity

Time Complexity: Color space conversion

O(n)

Understanding Time Complexity

When converting colors from one space to another, the time taken depends on how many pixels we process.

We want to know how the work grows as the image size grows.

Scenario Under Consideration

Analyze the time complexity of the following code snippet.

function rgbImage = ycbcr2rgb(ycbcrImage)
  [rows, cols, ~] = size(ycbcrImage);
  rgbImage = zeros(rows, cols, 3, 'uint8');
  for r = 1:rows
    for c = 1:cols
      Y = double(ycbcrImage(r,c,1));
      Cb = double(ycbcrImage(r,c,2)) - 128;
      Cr = double(ycbcrImage(r,c,3)) - 128;
      R = Y + 1.402 * Cr;
      G = Y - 0.344136 * Cb - 0.714136 * Cr;
      B = Y + 1.772 * Cb;
      rgbImage(r,c,1) = uint8(min(max(R,0),255));
      rgbImage(r,c,2) = uint8(min(max(G,0),255));
      rgbImage(r,c,3) = uint8(min(max(B,0),255));
    end
  end
end
    

This code converts an image from YCbCr color space to RGB by processing each pixel one by one.

Identify Repeating Operations

Identify the loops, recursion, array traversals that repeat.

• Primary operation: Nested loops over all pixels in the image.
• How many times: The inner code runs once for each pixel, so rows x columns times.

How Execution Grows With Input

As the image gets bigger, the number of pixels grows, so the work grows too.

Input Size (pixels)	Approx. Operations
10 x 10 = 100	About 100 times the pixel work
100 x 100 = 10,000	About 10,000 times the pixel work
1000 x 1000 = 1,000,000	About 1,000,000 times the pixel work

Pattern observation: The work grows directly with the number of pixels, so doubling width and height quadruples the work.

Final Time Complexity

Time Complexity: O(n)

This means the time grows in a straight line with the number of pixels in the image.

Common Mistake

[X] Wrong: "The time depends only on image width or height, not both."

[OK] Correct: The code processes every pixel, so both width and height matter together, multiplying the total work.

Interview Connect

Understanding how image size affects processing time helps you explain performance in real projects involving pictures or videos.

Self-Check

"What if we vectorized the code to process all pixels at once? How would the time complexity change?"

Practice

1. What does the MATLAB function rgb2gray do?

easy

A. Converts an image from CMYK to RGB

B. Converts a grayscale image to RGB

C. Converts an image from HSV to RGB

D. Converts a color image from RGB to grayscale

Solution

1. Step 1: Understand the function name

   The function name rgb2gray suggests conversion from RGB to gray scale.

2. Step 2: Recall MATLAB image processing basics

   In MATLAB, rgb2gray converts a color image (RGB) to a grayscale image by removing color information.

3. Final Answer:

   Converts a color image from RGB to grayscale -> Option D

4. Quick Check:

   rgb2gray = RGB to grayscale [OK]

Hint: Look for 'rgb2gray' meaning RGB to gray conversion [OK]

Common Mistakes:

• Confusing rgb2gray with gray2rgb
• Thinking it converts HSV images
• Assuming it changes image size

2. Which of the following is the correct syntax to convert an RGB image stored in variable img to HSV in MATLAB?

easy

A. hsvImage = rgb2hsv(img);

B. hsvImage = hsv2rgb(img);

C. hsvImage = rgb2gray(img);

D. hsvImage = gray2hsv(img);

Solution

1. Step 1: Identify the function for RGB to HSV conversion

   MATLAB uses rgb2hsv to convert RGB images to HSV color space.

2. Step 2: Check the syntax correctness

   The syntax hsvImage = rgb2hsv(img); correctly calls the function with the image variable.

3. Final Answer:

   hsvImage = rgb2hsv(img); -> Option A

4. Quick Check:

   rgb2hsv syntax = correct [OK]

Hint: Remember rgb2hsv converts RGB to HSV [OK]

Common Mistakes:

• Using hsv2rgb instead of rgb2hsv
• Using rgb2gray for color conversion
• Using non-existent gray2hsv function

3. What will be the output size of the variable grayImage after running this code?

rgbImage = imread('peppers.png');
grayImage = rgb2gray(rgbImage);

medium

A. Same height, width, and channels as rgbImage

B. Same height and width as rgbImage but single channel

C. Half the height and width of rgbImage

D. Double the height and width of rgbImage

Solution

1. Step 1: Understand input image dimensions

   RGB images have height x width x 3 channels (for red, green, blue).

2. Step 2: Effect of rgb2gray on image size

   rgb2gray converts to grayscale, keeping height and width but reducing channels to 1.

3. Final Answer:

   Same height and width as rgbImage but single channel -> Option B

4. Quick Check:

   rgb2gray output = height x width x 1 [OK]

Hint: Grayscale keeps size but reduces channels to 1 [OK]

Common Mistakes:

• Assuming rgb2gray changes image dimensions
• Thinking output has 3 channels
• Confusing with resizing functions

4. Identify the error in this MATLAB code snippet for converting an RGB image to grayscale:

img = imread('image.jpg');
grayImg = rgb2gray(img, 'double');

medium

A. imread cannot read 'image.jpg'

B. Variable names are invalid in MATLAB

C. rgb2gray does not accept a second argument like 'double'

D. Missing semicolon after imread

Solution

1. Step 1: Check rgb2gray function usage

   rgb2gray only takes one input argument: the RGB image.

2. Step 2: Analyze the second argument 'double'

   Passing 'double' as a second argument causes an error because rgb2gray does not support it.

3. Final Answer:

   rgb2gray does not accept a second argument like 'double' -> Option C

4. Quick Check:

   rgb2gray single input only [OK]

Hint: Check function docs for allowed arguments [OK]

Common Mistakes:

• Adding unsupported extra arguments
• Assuming imread fails without reason
• Thinking semicolon is mandatory for correctness

5. You want to convert an RGB image to HSV and then extract only the Hue channel for further processing. Which MATLAB code snippet correctly does this?

hard

A. hsvImage = rgb2hsv(rgbImage); hueChannel = hsvImage(:,:,1);

B. hsvImage = rgb2gray(rgbImage); hueChannel = hsvImage(:,:,1);

C. hsvImage = hsv2rgb(rgbImage); hueChannel = hsvImage(:,:,2);

D. hsvImage = rgb2hsv(rgbImage); hueChannel = hsvImage(:,:,3);

Solution

1. Step 1: Convert RGB to HSV correctly

   Use rgb2hsv to convert the RGB image to HSV color space.

2. Step 2: Extract the Hue channel

   In HSV, the first channel (:,:,1) is the Hue component.

3. Final Answer:

   hsvImage = rgb2hsv(rgbImage); hueChannel = hsvImage(:,:,1); -> Option A

4. Quick Check:

   Hue channel is first in HSV [OK]

Hint: Hue is channel 1 after rgb2hsv conversion [OK]

Common Mistakes:

• Using rgb2gray instead of rgb2hsv
• Extracting wrong channel index
• Using hsv2rgb instead of rgb2hsv